Dynamic Analysis of Participatory Learning in Linked Open Data: Certainty and Adaptation

Abstract

Graph-based data formats are popular ways of representing information, while graph-processing engines and graph databases become preferable tools for handling data of different size. World Wide Web Consortium has introduced a graph-based data format called Resource Description Framework (RDF) as the part of its Semantic Web initiative. The intrinsic features of RDF, i.e., its interconnectivity and simplicity of expressing information as triples containing two entities connected by a property, provide new possibilities of analyzing and absorbing information.

The participatory learning of propositional knowledge is an attractive way of integrating and updating knowledge bases built based on symbolic data equipped with uncertainty. In such context, an idea of considering RDF triples as propositions allowed us to use the principles of participatory learning for assimilating RDF triples and handling different levels of uncertainty associated with them.

The paper examines the RDF-based participatory learning process from the perspective of its dynamics. The emphasis is put on aspects related to handling certainty, accepting new pieces of information, and dealing with contradicting information. The learning process is presented, and the results of analysis are provided.